Aalborg Universitet. Specification of a Two-Dimensional Test Case Nielsen, Peter Vilhelm. Publication date: 1990

Transcription

1 Aalborg Universitet Specification of a Two-Dimensional Test Case Nielsen, Peter Vilhelm Publication date: 199 Document Version Publisher's PDF, also known as Version of record Link to publication from Aalborg University Citation for published version (APA): Nielsen, P. V. (199). Specification of a Two-Dimensional Test Case: (EA). Aalborg: nstitut for Bygningsteknik, Aalborg Universitet. Gul Serie, No. 8, Vol.. R94 General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.? Users may download and print one copy of any publication from the public portal for the purpose of private study or research.? You may not further distribute the material or use it for any profit-making activity or commercial gain? You may freely distribute the URL identifying the publication in the public portal? Take down policy f you believe that this document breaches copyright please contact us at vbn@aub.aau.dk providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from vbn.aau.dk on: september 27, 218

2 NSTTUTTET FOR BYGNNGSTEKNK DEPT. OF BULDNG TECHNOLOGY AND STRUCTURAL ENGNEERNG AALBORG UNVERSTETSCENTER AUC AALBORG DANMARK nternational Energy Agency, Energy Conservation in Buildings and Community Systems, Annex 2: Air Flow Pattern Within Buildings PETER V. NELSEN SPECFCATON OF A TWO-DMENSONAL TEST CASE NOVEMBER 199 SSN R94

3 Printed at Aalborg University

4 NSTTUTTET FOR BYGNNGSTEKNK DEPT. OF BULDNG TECHNOLOGY AND STRUCTURAL ENGNEERNG AALBORG UNVERSTETSCENTER AUC AALBORG DANMARK nternational Energy Agency, Energy Conservation in Buildings and Community Systems, Annex 2: Air Flow Pattern Within Buildings PETER V. NELSEN SPECFCATON OF A TWO-DMENSONAL TEST CASE NOVEMBER 199 SSN R94

5

6 RESEARCH TEM NO SPECFCATON OF A TWO-DMENSONAL TEST CASE Peter V. Nielsen, The University of Aalborg, Denmark NTRODUCTON This paper describes the geometry and other boundary conditions for a test case which can be used to test different two-dimensional CFD codes in the lea Annex 2 work. The given supply opening is large compared with practical openings. Therefore, this geometry will reduce the need for a high number of grid points in the wall jet reg1n. The geometry is suitable for test of different CFD codes because there exist a number of measurements of the flow. The geometry has also been used ealier for test of computer codes and could therefore be standardized as a benchmark test for room air distribution models. SPECFCATON OF SOTHERMAL TEST CASE (case 2Dl) Figure 1. Model for test case and experiments. System of co-ordinates.

7 2 The geometry of the test case has the following specification, see figure 1. L/H = 3., h/h =.56 and tfh =.16 (1) or H = 3. m, L = 9. m, h =.168 m and t =.48 m (2) Obviously the height of the supply opening is rather large compared with practical diffusers. This geometry will reduce the need for a high number of grid points and it is possible to make an inlet specification direct at the opening instead of using the box method or the prescribed velocity method. A large opening will also ensure that the main part of the inlet flow passes outside the boundary region of the ceiling where different boundary conditions are used in CFD codes. The inlet conditions for the velocity are given by h u Re = -- = 5 ll (3) or U =,455m/s (4) for a kinematic viscosity of v = 15,3 1-6 m 2 /sat an inlet temperature of 2 C. The Reynolds number is based on the slot height because the flow in the ceiling region and in the rest of the room is strongly influenced by the inlet conditions. nlet conditions for the turbulent kinetic energy k and dissipation E are given by ko = 1.5 (.4 u ) 2 (5) (6) where f = h/1 (7) The inlet conditions correspond to a turbulence intensity of 4%. t is difficult to stipulate but variation of the length scale fo within a reasonable level shows only a very small influence on the flow in the room. The production and dissipation of turbulence in the shear layer in the downstream flow will dominate the possible effect of different inlet values, see [1].

8 3 Prediction of the flow should be given at the two vertical lines X= H' X= 2H (8) and at the two horizontal lines y = h/2, y = H- h/2 (9) Figures 5, 6, 7, 8 and 1 show measurements at those locations which can be used for comparison. The isothermal test case (2D1) can be extended by a mass transport equation and constant mass flux along the floor. (This situation may be simulated by an energy equation and constant heat flux along the floor for the Archimedes number Ar The buoyancy term in the Navier-Stokes equation had to be ignored). Prediction of the concentration profile should be given at the horizontal level y =.75H (1) Figure 11 shows measurements at this location which can be used for comparison. The measurements were made at a Reynolds number of 71. The literature shows two dimensional predictions in the selected geometry. Predictions in reference [2] are based on the numerical solution of the vorticity and stream function equations and predictions in references [1], [3] and [4] are based on numerical solution of the continuity and momentum equations. SPECFCATON OF NON-SOTHERMAL TEST CASE (case 2D2) The aim of this test case is to predict flow with a strong buoyant effect. A constant heat flux is added along the floor and the Archimedes number is raised until the CFD code predicts a flow with a reduced penetration depth Xre The Archimedes number Ar is defined as Ar = _(3_ g_ h_ _b._t_o u2 (11) where (3, g and b.t are volume expansion coefficient, gravitational acceleration and temperature difference between return and supply, respectively. The prediction will be made in the same geometry and at the same velocity level as used in the isothermal case (1), (2), (3) and (4).

9 4 Xre/H X re s w a Ar Figure 2. Penetration depths Xre in models with different inlet geometry. Figure 2 shows the results of a number of model experiments made by Schwenke [5]. t is indicated that a cold supply jet will deflect from the ceiling before it reaches the end wall for a Archimedes number Ar equal to or larger than.2. t should be stressed that experiments for h/ H =.56 and Re= 5 are missing. A selection of Ar =.2 corresponds to a temperature difference between return and supply of (12) in the geometry with room height of 3 m and a supply velocity of.455 m/s. The low velocity in the room ensures a buoyant flow at the given level of temperature differences. The heat flux is calculated from 6.T, u and h. The predictions are repeated for increasing Archimedes' number until a reduced penetration depth Xre takes place Ar = (13) The length of the penetration depth Xre is in some cases dependent on the initial conditions. This is shown in reference [5] where two different Xre are obtained when the same measuring conditions are reached through increasing Archimedes' number and decreasing Archimedes' number. This effect can be predicted with a CFD code as shown in reference [6]. The flow field should be given in the same way as in the isothermal case, see (8) and (9).

10 Furthermore, a horizontal section showing the part of the jet should be given for the Archimedes number where reduced penetration takes place. A depiction of the maximum velocity Urm in the occupied zone, as a function of the different Archimedes' number, will also be relevant, see figure 3 in reference [6]. LASER-DOPPLER MEASUREMENTS The measurements described in this chapter are made by laser-doppler anemometry in a model with the geometrical specification of (1) and with WH = 1. where W is the width of the model. The model has a real size of H = 89,3 mm. The work is a part of A. Restivo's Ph.D. thesis [1]. Laser-doppler anemometry is suitable for measuring the velocity in recirculating flow compared with hot-wire and hot sphere equipment. Figure 3 shows the u-velocity and the intensity vf:j2 in four different points as a function of the Reynolds number. The flow seems to be fully turbulent for a Reynolds number of 5. Figure 4 shows the inlet profiles. The profiles are unfortunately slightly influenced by the inlet geometry which has a contraction in the z-direction and constant height. Measurements of the rms velocity justified the value of 4% in the predictions, see (5). Figures 5 and 6 show the mean and rms velocity at the two vertical positions x H = 1. and 2.. The upper part of the flow is typical for a wall jet. The figures show that the main part of the flow has a high turbulent level compared with the mean velocity which means that measurements with hot-wire anemometry would be very difficult. The turbulent intensity vf:j2 is the upper and lower part of the model at xl H = 2. is typical of a wall jet. The intensities in the two other directions in a twodimensional wall jet may be given as v' 2 "'.6 u' 2 and w' 2 "'.8 u' 2. This means that the turbulent kinetic energy has a level rather close to the intensity vf:j2 5 (14) Equation (14) may be used for a rough validation ofthe predicted turbulent kinetic energy. The turbulence intensity vf:j2 in the lower part of the model at x H = 1. is rather close to the mean velocity u which means that the turbulence is transported to this area. Comparisons between mean velocity in figures 5 and 6 show that the flow is slightly three-dimensional. This is also recorded with hot-wire measurements g1ven m figure 1 for W H =.5 and 1.. Figures 7 and 8 show u and vf:j2 close to the top and bottom wall. t should especially be observed that recirculating flow takes place in the corner at the wall opposite the inlet as well as in the corner below the inlet.

11 6 u/u P./uo o u/u.65 (H. h, ) l:;.,f:jiluo Re f ~ H, h,.4 W) Re [:;. 6. [:;. [:;. [:; t- (2H, H-h, ) Re [:; [:;. 1- (2H, H-h,.4W) Re -.1 Figure 3. Dependence on mean and rms velocity on Reynolds' number, see [1].

12 u/u z/w... x/h y = h/ u/u... Ptuo /:?-/uo u/u y/h x/h =. z/w =..4 C> o u/u C>... eo,p/uo... C>.25 C>... eo C>... eo C>... eo C>....5 C>... eo C>... eo C>... eo C>... eo.75 C>... «:> C>... ~ C>... ~ [::2/uo Figure 4. nlet pro les, Re = 5, see [1].

13 8 Figure 5. Measured mean and rms velocity in the symmetry plane, Re see (1]. 5,

14 9 -.:t d 11 ~ ::J -.:t N -- ::J d ::J ~ N d 66 6 d q q -- d X N Q) N c:i tp -.:t d -.:t c:i N c:i q 11 d x -- ::J ::J ::J ~ 6 Figure 6. Measured mean and rms velocity in the plane z /W =.4, Re = 5, see [1).

15 1 u/u, {Ji.;uo y:h/2 1. z/w:o.o U/U -{:;2/uo ~ u/uo fji./uo.5 6J x/h Y= H-h/2 z/w=o.o &,~66 d ::,.66 6 o Ooo oo x/h Figure 7. Mean and rms velocity near the top and bottom walls in the symmetry plane, Re = 5, see [1].

16 11 u/u {;2./uo 1. y = h/2 z/w=.4 u/u -{;2./uo , u/u P/uo x/h y = H-h/2 z/w= x/h Figure 8. Mean and rms velocity near the top and bottom walls at z/w =.4, Re = 5, see [1].

17 12 OTHER MEASUREMENTS The measurements described in this chapter are made by hot-wire anemometry, pitot tube and thermocouples in a model with the following geometrical specification, see [2] h/h =.56, L/H = 3., tjh =.16 W/H = 4.7, 1. and.5 (15) The model has a real size of H = 127 mm. z/w y/h = f- y/h. 1. y/h. 1. y/h z/w =.33 z!w=.. z/w = L'-...J 1. u/u Figure 9. Velocity distribution at the supply opening measured by pitot tube, R e = 5, W/H = 4.7.

18 The velocity distribution in the supply opening is shown in figure 9. The velocity is rather constant across the whole area of the slot. 13 y/h. r , ~ , o W/H = 4.7 D W/H = 1. \7 W/H = u/u Figure 1. Hot-wire measurements of mean velocity at x/ H = 2. and z/w = for three different widths of the model, Re = 71, see (2]. Figure 1 shows the influence of the width W. The measurements suggest threedimensional flow but it is not possible to exclude some influence from the probe support due to the small width of the model. t is difficult to compare the results with measurements in the last chapter because the inlet profiles are slightly different in the two cases. Figure 1 shows a large area in the middle of the flow where hot-wire measurements of mean velocity are impossible.

19 14 T- T Ll T r- Re =71 Ar = 3.1 X 1-6 o Ar =1.8x \7 Ar = 1.1 x r- 8' D 1.4 r xl H Figure 11. Measured temperature distribution at y / H =. 75 and z /W =.17 in model with heat flux through the bottom surface, Re= 71, W/H = 4.7, see [2]. Tw- To Ll T r- Re =71 o Ar=3.1x1-6 o Ar = 1.8 x \7 Ar = 1.1 x 1 1r- s \7 5 r- OL-----~---~~----~'----~~-----~~---~ x/h Figure 12. Measured temperatures T w at the bottom surface in the experiment shown in figure 11.

20 Figures 11 and 12 show the air temperature and surface temperature measured in an experiment with heat flux through the bottom surface. The Archimedes numbers in the experiments are small and the velocity distribution will therefore be similar to the measurements shown in figure 1. t is difficult to produce a constant heat flux in the experiment but comparisons between predictions with different heat flux distributions at the bottom surface show an insignificant influence on the temperature profile at y / H = REFERENCES [1] Restivo, A., Turbulent Flow in Ventilated Rooms, Ph.D. thesis, University of London, March [2] Nielsen, P.V., Flow in Air Conditioned Rooms (English translation of Ph.D. thesis, Technical University of Denmark, 1974), Danfoss A/S, Denmark, [3] Nielsen, P.V., A. Restivo and J.H. Whitelaw, The Velocity Characteristics of Ventilated Rooms, Journal of Fluids Engineering, September 1978, Vol. 1, pp [4] Davidson, L. and E. Olsson, Calculation of some Parabolic and Elliptic Flows using a new One-Equation Turbulence Model, Proc. 5th nternational Conference on Numerical Methods in Laminar and Turbulent Flow, Montreal, Vol. 1, [5] Schwenke, H., Uber das Verhalten ebener horizontaler Zuluftstrahlen im begrenzten Raum, Luft- und KaJ.tetechnik, No. 5, Dresden, [6] Nielsen, P.V., A. Restivo an J.H. Whitelaw, Buoyancy-Affected Flows in Ventilated Rooms, Numerical Heat Transfer, Vol. 2, pp , 1979.

21

22

23